Funding, Sustainability and Governance
From GRDI2020
This is a GRDI challenge; return to Main Page with all the challenges and recommendations
Contents |
Introduction
Basic funding dynamics of individual service
Already in the case of an individual data infrastructure the funding, sustainability and governance issues are interlinked and form quite a complex system. For example, the amount of funding, the length of time it can be guaranteed in the future and the constraints the funding agenda imposes on the governance models (and use of the data infrastructure) will influence sustainability both directly (length of the funding cycle) and indirectly (satisfaction of the existing users and the ability of the data infrastructure to convince new user communities to join in as users). On the other hand, the number of users tends to improve the funding situation.
Emerging global scale dynamics
In a global research data infrastructure these dynamics of the individual data infrastructure components form an even more complex feedback system. Interoperability of the data infrastructure components may make predicting the load of the system more difficult. The governance model becomes more complex if there is a wish to accommodate “remote” users. These developments may in turn make the “local” users less satisfied with the system, which influences the funding and sustainability of the component negatively.
Current project-based funding models, that make funding decisions mainly based on numbers of users and user communities, may create situations where it is difficult to incentivise individual actors to work towards interoperability. “Remote” users may be difficult to account for and report reliably to the funding agency, while the interoperability may also allow current users to migrate to other solutions. The unintended consequence of the current de facto impact assessment methodologies is the creation of somewhat vertically integrated solutions.
Emergent collaborative data management as competitor
One of the ways these vertically integrated solutions may emerge is from the communities consisting of closely collaborating groups of people who could be characterised as “Terabytes of data on the desktop” groups. For these groups data management is becoming a moderately pressing concern, but the most natural solutions may emerge from e.g. few of these groups looking into solving issues related to backing up of the data in collaboration, perhaps through some kind of encrypted peer-to-peer storage system. Once these systems are put in place, the natural next steps would be to look into controlled sharing of the data that is already physically stored in a distributed manner.
This kind of “emergent collaborative data management” will remain a strong competitor for the GRDIs – especially if they don’t provide a low-level interface that looks like a disk mounted over the network. Sharing of very basic resources between collaborating parties leverages the natural social capital (trust) between the participants and will not require dealing with external parties. Thus the changes to the collaboration practices between the parties will remain minimal.
Bootstrap options
Attracting the users (and potential users) from these kinds of ad hoc data management solutions poses certain challenges in terms of funding and sustainability of the GRDIs. Providing low-level service interface to the data storage system puts the GRDI in direct competition with commercial Cloud/utility computing providers, who are much better placed to utilise the various economies of scale. Couple with the anticipated need to address the underlying basic infrastructure issues in terms of intra- and inter-campus network and computing infrastructure (the general “Digital divide” question) may make it very challenging to report positive impact to the funding agencies in short term: investments in the infrastructure and GRDI may seem to just push more users towards the commercially available solutions.
Governance and sense of ownership
To alleviate this issue, the governance models need to maximise the sense of ownership for all critical stakeholders in the GRDIs: funding agencies, end-user application developers, users organised into groups and individual researchers just starting with data management activities. The GRDI service provider community may also need to manage the expectations of the funding agencies carefully.
10-year vision
It is difficult to develop a vision for governance and funding without a clear idea on how the Data Infrastructures ecosystem will evolve. By 2020 it is envisaged that a hybrid distributed ecosystem of disciplinary data infrastructures (e.g. HEP, Bio) and multi-disciplinary ones (e.g. in new areas such as Arts and Humanities) will co-exist and experts will be striving to make this ecosystem interoperable so as to be able to reuse best practices and promote effective inter-disciplinary collaboration, thus promoting interoperability, data exchange, data preservation and distributed access among others. Therefore, separate governance for each of the ecosystem building blocks will still exist, however with possible collaboration mechanisms among disciplines (e.g. observers of other disciplines in disciplinary bodies) with global inter-disciplinary governance bodies shaping up discussing collaboration and interoperation strategies. Collaboration or possible disciplinary or multi-disciplinary integration with other e-Infrastructure components (such as the computing that will process the data) may appear.
Challenges and Recommendations
- Link data management and grant process
- Collate, categorise and learn from best practices for sustainable data infrastructures
- Create a multi-tiered framework for certification of repositories
A related recommendation with regard to Data Storage and Data Curation and Preservation is to Define reference SLAs to cluster data retention requirements.
